bin/map: Deal with fields that may be in the dataset under more than one prefix by getting all fields and coalesce()ing them (e.g. SpeciesLink has dwcore* and darwin1* columns for the same DwC field)